Abstract:
|
Basket trial design is a recent practice in the clinical trial field since it could not only examine one treatment on several related diseases in a single trial but also examine a biomarker response across multiple diseases. The number of patients with a putative biomarker within a single disease is small, which makes it difficult to enroll adequate number of patients, and basket trial will pool the responses from the same biomarker. The sequential conditional probability ratio test (SCPRT) is one of the commonly used group sequential designs. It can control the discordant probability, which stands for the chance that get reverse conclusion from interim analysis to non-sequential analysis. Existing such design is for one-dimensional response. In this project, our aim is to obtain the generate boundaries of discordance probability and type I errors for different stages under overall type I error are calculated and the boundaries for each stages and diseases for a SCPRT basket trial. Thus, we have proposed an SCPRT basket design with its sequential boundary modulated by a discordance probability. The proposed procedure is applicable to most clinical trials and since it accrues patien
|